Advertisements or Answers

I struggle regularly with the visions of our company. Are we a software company or a media company (licenses or ads) or can we really be both? I am still on board with Microsoft as a software company and question whether advertisements really represent huge opportunities when not directly associated with content (such as TV shows, news articles, etc.). But what if that content was your own? I am referring specifically to a recent announcment, found on Brand Week, that talks about how Microsoft will be able to scan email and use the information it finds to target search queries and advertisements, right to the desktop.

What I see here is an opportunity far beyond advertisements, one of answers. Imagine a similar agent actually running on the desktop, reading what you're reading and typing, and actively displaying (in a non-obtrusive manner) related content on the web, in news groups, on your desktop, in your email archives, and on corporate web sites and file shares. Even better, the agent remembers what you click-thru and learns your preferences for content.

Communication is a huge problem. We thought we had it solved with email; just send your announcement in email and you can be sure people got it. When people stopped reading their email and complaining about the fact they weren't regularly informed, we moved to portals. This actually shifted the process from a push-based communication to a pull-based communication. It was now possible for people to learn what the needed to know, when they were prepared (read: interested) to read it. Unfortunately, everyone jumped on the portal bandwagon and now we have so many of them, no one can find what they need.

So, along comes search, which allows us to find what we know we want (albeit still a bit sloppily), but fails when it comes to informing us of what we don't know we want. Here is where advertising comes along. In the aforementioned system, imagine that anyone in a large organization could actually create advertisements for information that might have previously been disseminated in email. Those "ads" could be linked to keywords and corporate user profiles (market targets) and appear on the desktops of fellow employees at the time when they are most likely to read them. Let me provide an example.

Internal Microsoft systems know much about me. They know my group, my role, my experiences (projects), my areas of specialty, etc. As a consultant with regular customer touch on infrastructure products, the Exchange team might decide that I should know about a new Rapid Deployment Program, but the person running the program has never even heard of me or the 1000s of others out there like me. However, they do know that they want their announcment of the program to reach sales and services personnel who've worked on Exchange projects. They build their "ad/announcment," publish it to an ad-center like server, and it pops up on my desktop. The timing of the pop-up is what is critical. It should happen when I am reading a document about Exchange, sending an email asking a question about Exchange, or visiting an Exchange-specific web site. It should not popup when I am researching or working on Windows, because I am not likely to be interested in reading it at that time (I tend to focus on one thing at a time). Also, it should be clean and not overburden me (highly focused target, limited to 3 ads/announcement maximum displayed). Finally, it should be flexbile. If I am reading a really large white paper with mutliple chapters, it should know which chapter I am current reading (by what is displayed on the screen) and focus the results on that chapter.

I know that some people don't like the idea of an agent running on your machine and observing everything you do and maybe this idea isn't for everyone. However, large organizations suffer greatly from communications problems and the most challenging part of the problem is tying a recipient's interests with a publisher's agenda; something that email, portals and search fail at, but targeted advertising makes quite possible. 

Internally at Microsoft, we still receive (and don't read) hundreds of email messages per day and many of us refer to these messages as internal spam. The stems from the fact that too many people are trying to connect with too many others without a clue of who or how to target their messages. I contend that if these messages were targeted as advertising, we might start reading more and might feel better connected with those whom we don't even know.

I will cite two examples of how this idea already works, in a less than fully baked solution. Amazon.com, a place I truly enjoy shopping, is first. The more I purchase from them, the better my shopping experience gets. 60% of the advertisements/recommendations that show up on the front page of Amazon are things I am interested in buying (if money and storage were unlimited). And almost half of that 60% is for things I didn't even know I wanted. To me, that is some effective communication. When was the last time you watched a TV show and wanted 60% of the items advertised? When was the last time you read 60% of the email not specifically addressed to you (ie: sent to a distribution group that you belong to)? When was the last time 60% of the search results were relevant to your question or problem?

The other example is the old Napster. The interface itself was pretty limited, but the user community around it adapted a very similar model to what I am prosposing in this post. Because only artist, title and file name were actually searchable in the tool, content providers began altering these values so that searchers would discover them. For example, someone searching for Grateful Dead might have been trying to download album releases of the music, but might have accidentally discovered the whole spectrum of live recordings that are shared among enthusiasts. This person might not have even known such recordings existed. Once they learned to search for "LIVE" instead of a song name, their world got bigger. It has always amazed me how most people viewed Napster 1.0 as a way to get free music (and most of it illegally, I might add). However, from a software development perspective, Napster was probably the MOST collaborative platform ever developed and would have been far more useful if its content was the millions of documents, white papers, and email conversations currently locked up on individual computers in Corporate America.

Microsoft is showing some potential in the advertising arena, but I believe the real value in these products is what we might someday to with them to achieve better communication in a large organization.